Led by Dr. Daphne Koller, together with Drs. Robert Tibshirani and AtuI Butte, this series of bioinformatics projects will focus on the development of computational methods for analyzing the cellular network underlying the immune system and for understanding how this network is perturbed by the response to vaccine. This project will build on our successful development of new deconvolution software (csSAM) to extend it to the development of applications to other datasets in addition to whole blood gene expression such as serum cytokines and phosphoflow. Building on existing successful tools this team has previously developed, cell-specific deconvolved data will then be used to construct the regulatory network of each of the different cell-types present in peripheral blood as well as the inter-cellular network between them. Integrating the data collected from all ofthe other projects in this proposal, this project will develop computational tools for predicting vaccine responses, and for identifying common vaccine mechanisms and genetic correlates with response. The learned models will provide novel scientific insights regarding the immune system, including: the relationships between genotype, gene expression, phosphoprotein levels, and phenotype; the interactions between genes within an immune cell; and the cytokine-mediated interaction between different immune cells. All these will be used for the specification of human immune metrics. The specific aims for Project 7 are to develop: Aim 1: Methods for analysis of data derived from mixed samples; Aim 2: Methods for using diverse data to construct network models for immune system cells; Aim 3: Methods for understanding the factors affecting immune response; and to: Aim 4: Use systems-level analysis to identify causal polymorphisms that affect immune response.